Using a Cognitive/Metacognitive Task Model to Analyze Students Learning Behaviors

Adapting to learners’ needs and providing useful, individualized feedback to help them succeed has been a hallmark of most intelligent tutoring systems. More recently, to promote deep learning and critical thinking skills in STEM disciplines, researchers have begun developing open-ended learning environments that present learners with complex problems and a set of tools for learning and problem solving. To be successful in such environments, learners must employ a variety of cognitive skills and metacognitive strategies. This paper discusses a framework that combines a theory-driven, top-down approach with a bottom-up, pattern-discovery approach for analyzing learning activity data in these environments. Combining these approaches allows for more complex qualitative and quantitative interpretation of a student’s cognitive and metacognitive abilities. The results of this analysis provide a foundation for developing performanceand behavior-based learner models in conjunction with adaptive scaffolding mechanisms to promote effective, personalized learning